MATERIAL Y MÉTODOS
SIGNATURA DE LAS FIGURAS
Robots do not emerge from engineering labs as sleek neutral entities, ready to impose on an unsuspecting labour force. Beyond their technical capabilities, the roles and the ways that they interact with humans are carefully engineered into their very being – and much of this design is framed, both intentionally and unintentionally, around stereotypical cultural norms about who ‘men’ and ‘women’ are and what they do. Human–Robot Interactions (HRI) research explores what types of interactions robots are intended to have with humans. A starting point for these investigations is that to function with maximum efficiency,
robots must take on certain characteristics which allow people to fill in the gaps and build a mental model
of who the robot is – what does it do? How will it respond? This is described as Computers as Social
Actors theory, in which humans ascribe agency and personality to
computer–mediated interactions – in other words, that humans
instinctively treat computers like humans. But which humans? If
a robot is intended to induce affection, it will be designed one
way; if fear, another – even if the core purpose of the machine remains the same. Roboticists
are advised that ‘companion robots’ for children should not be designed to carry features which are likely to frighten them, such as an overwhelmingly large size or “creepy appearance”.
What induces fear, affection or trust in humans is, however, loaded with user expectations, which in turn are embedded in a variety of culturally–specific norms and stereotypes. People quickly form mental models based on what they know best, namely other people: age, race and gender provide the most salient social cues, and are deeply fundamental parts of how people understand and respond to each other. If a robot is expected to embody female traits, how will it be designed? And what expectations are being collected under the category of ‘female’?
Physical traits are obvious ways of denoting gender, although the ways in which this has played out in robot design scenarios appear to have been dropped in from the 1950s. In a field which continues to be male–dominated, gender is often approached using what Jennifer Robinson dryly describes as “common sense knowledge…of femininity in relationship to masculinity (and vice versa)”. ‘Female’ robots have been so denoted through pink lips, long hair and a higher pitched voice, and by name and gender verbal introduction. Robot designer Tomotaka Takahashi complained of the difficulties of engineering his Female Type robot because he argued that the feminised version would have to have a slimmer torso, which presented a challenge in terms of fitting the internal mechanisms into the body. Finally, gender can also be prompted by occupational identity: even if it has a gender–neutral appearance, users will still assign a gender based on what tasks the robot is doing.
Once robots have been cued as male or female, user expectations will fill in the gaps with men and women particularly likely to
anthropomorphise voices that correspond to their own gender identity. ‘Female’ robots are seen as more likeable, trustworthy, persuasive and communal; whilst ‘male’ robots are perceived as being more threatening and agentic, although the praise they give is taken more seriously. Men also tend to feel psychologically close to robots that have a ‘male’ voice, and are less willing to take instructions from the car’s SatNav if it has a ‘female’ voice.
As “a robot may be mistakenly viewed as medically competent if it is dressed in a lab coat and wearing a stethoscope”, so gendered robots have certain skills ascribed to them. ‘Male’ robots are assumed to be better at repairing technical devices whilst ‘female’
Part 4. Robots and Justice
robots are assumed to be more suited to domestic and caring services. Roboticists themselves have fallen in line with these lines of thought. In their 2005 paper, eliciting information from people with a gendered humanoid robot, members of the Human– Computer Interaction Institute at Carnegie Mellon University built design scenarios based on the notion that “women are more knowledgeable about dating forms and social practices, and they have more social skill than men do”.
What is worrying is how these reactions are blandly, neutrally, co–opted into design processes simply as a way of inducing a certain type of user reaction. Gender norms are not challenged, but instead reproduced in the name of machine utility. Security robots which are coded as ‘male’ are deemed more useful – and thus acceptable – than with a ‘mismatched’ identity. Male roboticists predict that their female robots will find work in classic sites of affective labour in the service sector: in bars, information booths and ‘upmarket coffee shops’. Some have contended that creating robots which appear distinctly ‘male’ or ‘female’ is preferable in certain contexts:
A mechanic’s helper robot, if stereotypical, would be male. If we wanted this robot to have minimal and efficient conversation with users about what tools they need, how to assist, and so forth, then the mechanic’s helper should be male. Suppose that we wanted users to provide more information, to explain themselves, to be redundant. This might be a design goal if the robot was not a mechanic’s helper robot, but a general assistant and not specifically designed for the task. Then, the robotic assistant should be anti–sterotypic for the task, ie. Female for the mechanic’s helper.
Members of the robotics community have raised concerns that their peers are uncritically reproducing, reinforcing and perpetuating stereotypes attached to gender and work. Whilst some have described how developing ‘female’ assistive robots would be useful for facilitating bonding with infants, others have voiced their worries that children raised with these female robotic caregivers may then have their beliefs influenced about the nature of gender roles; for example that women are naturally nurturing and “derive immense joy from menial household chores”.